Code is manipulating LINQ results somehow?

I have a small application which is doing something that I cannot explain.  

I use a "LINQ Where" query  to identify some nodes in a tree.  This returns an iterator of Tree Nodes that match the where criteria. I delete those nodes and when I check the contents of the LINQ iterator, it contains nodes which were not part of the LINQ query.  

 

Specifically the application does the following:

• convert a list of folder paths to a simple tree structure
• use LINQ to select terminating paths (folders with no sub-folders)
• convert/copy the linq iterator to a List using "toList()"
  
• loop over the list and delete these nodes from the tree
• >> After processing, the LINQ iterator contains 2 nodes
  

The folder list resembles the following

         "\base",
         "\base\001",
         "\base\001\cfg",
         "\base\001\cfg\network",
         "\base\001\cfg\network\mgmnt", // terminating leaf
         "\base\001\cfg\network\users", // terminating leaf
         "\base\002",
         "\base\002\copy",
         "\base\002\copy\cfg",
         "\base\002\copy\cfg\network",
         "\base\002\copy\cfg\network\mgmnt", // terminating leaf
         "\base\002\copy\cfg\network\users", // terminating leaf

 

The LINQ iterator identifes 4 nodes (because these folders have no sub-folders)

         "\base\001\cfg\network\mgmnt"
         "\base\001\cfg\network\users"
         "\base\002\copy\cfg\network\mgmnt"
         "\base\002\copy\cfg\network\users"

 

If I inspect the LINQ iterator after deleting these 4 nodes from the tree, the iterator now contains 2 nodes

          "\base\002\copy\cfg\network"

          "\base\001\cfg\network" 

 

I would have expected the iterator to be empty.   

 

How does this happen???